true reading is therefore four degrees and four-tenths. This reading, however, must be corrected, because the original volume instead of being 200 cubic centimeters, is 235 cubic centimeters. The actual percentage of sugar in the sample examined is obtained by the following proportion: 200 : 235 = 4.4 : _x_. The correct reading is therefore 5°.2, the percentage of sugar in the sample examined. The results obtained by the method just described may vary somewhat from the true percentage by reason of the variation of the content of fiber in the bagasse. It is, however, sufficiently accurate for technical control in sugar factories and on account of its rapidity of execution is to be preferred for this purpose. More accurate results would be obtained by drying the bagasse, and proceeding with the examination in a manner entirely analogous to that described for the extraction of sugar from dried canes by aqueous alcohol. In both instances the reducing sugar is determined in the manner already mentioned. =215. Determination of Fiber in Cane.=—In estimating the content of sugar in canes by the analysis of the expressed juices, it is important to make frequent determinations of the fiber for the purpose of obtaining correct data for calculation. In periods of excessive drought, or when the canes are quite mature, the relative content of fiber is increased, while, on the other hand, in case of immature canes, or during excessive rainfalls, it is diminished. The chief difficulty in determining the content of fiber in canes is found in securing a representative sample. On account of the hard and fibrous nature of the envelope and of their nodular tissues, canes are reduced to a fine pulp with great difficulty by the apparatus in ordinary use. A fairly homogeneous pulp, however, may be obtained by means of the shredder described on page 9. The canes having been shredded as finely as possible, a weighed quantity is placed in any convenient extraction apparatus and thoroughly exhausted with hot water. The treatment with hot water should be continued until a few drops of the extract evaporated on a watch glass will leave no sensible residue. The residual fiber is dried to constant weight at the temperature of boiling water, cooled in a desiccator and rapidly weighed and the percentage of fiber calculated from the data obtained. On account of the great difficulty of securing a homogeneous pulp, even with the best shredding machines, the determination should be made in duplicate or triplicate and the mean of the results entered as the percentage of fiber. The term fiber as used in this sense, must not be confounded with the same term employed in the analysis of fodders and feeding stuffs. In the latter case the term is applied to the residue left after the successive treatment of the material with boiling, dilute acid and alkali. The analysis of canes for feeding purposes is conducted in the general manner hereinafter described for fodders.